Trades get organized by what’s painted on the truck. There’s a question that cuts across all of them and predicts more:
How easily can your people leave?
That single question determines whether the capacity limit you calculated is a number you can plan against or a number that moves under you. And it’s the difference between a business that can be scaled deliberately and one that has to be grown carefully.
Two groups
Low barrier to exit. Cleaning, appliance repair, handyman work. The skill is learned in months rather than years. The tools are affordable. No licence stands between an employee and their own business. So people leave to work for themselves — regularly, and usually your best ones, because the best ones have the most confidence and the most repeat customers willing to follow them.
High barrier to exit. HVAC, electrical, plumbing, larger construction. Licences take years and cost money. The equipment is capital. And in construction specifically, each specialist does one slice of a larger job — a framer cannot go and build a house alone, which is a structural barrier that has nothing to do with regulation.
Handyman work sits at the extreme low end, and it’s worth saying why: the entire proposition is one competent person with a van. There’s no team to leave, no specialization to lack, and no licence to acquire. Anyone good enough to be worth employing is good enough to work alone tomorrow.
Why this belongs in a marketing plan
Because your capacity limit is an input to every calculation about budget, and in one of these groups it’s stable and in the other it isn’t.
Run the two-point test in a licensed trade and Point B — full capacity — means something durable. You can plan a twelve-month scale-up against it, hire ahead of it, and expect the crew you have in March to be the crew you have in October.
Run the same test in a low-barrier trade and Point B is a snapshot. A third of your team may turn over inside the year, and every departure takes some capacity with it — sometimes along with the customers who preferred that particular person.
Which produces a rule that isn’t obvious and rarely gets written down:
In a low-barrier trade, a scale-up requires a hiring plan that starts before the campaigns do. In a high-barrier trade, the hiring plan can follow the demand.
Get that backwards and you produce the classic failure: leads arrive, the team can’t absorb them, work goes out late, reviews suffer, and the owner concludes the advertising was the problem.
Worth understanding what happens to the people who leave, since it affects both retention and the ones who return. Employed, the work was the work. Independent, they have to find the customer, sell the first visit, do the job, sell the next one — and then handle the software, the bookkeeping, the insurance and the legal side. A great many discover the job they left was worth more than it appeared. Some come back. That doesn’t stop the churn, but it explains why the churn has a floor.
One business, three sets of economics
Before the scaling story, a detour that matters more than the scaling story.
We were asked what an HVAC crew can handle in a week, and the honest answer is that the question doesn’t have one. HVAC isn’t a trade; it’s two or three trades sharing a truck. Service and repair — which always begins with a diagnostic visit — and installation. In practice they behave like separate departments, often with their own progression: a technician grows through service and moves toward installation, where the money is.
That’s not just an operational observation. Rounded figures from one company’s own model, by service line:
| Service line | Average job | Margin | Most you can pay per click | What the market charged | Verdict |
|---|---|---|---|---|---|
| HVAC service & repair | ~$750 | ~50% | ~$4.70 | ~$4.50 | Green — barely |
| HVAC installation | ~$11,000 | ~10% | ~$11.50 | ~$7 | Green, comfortably |
| Appliance repair | ~$300 | ~36% | ~$1.90 | ~$5.30 | Red |
Three things in that table are worth more than the rest of this article.
The margin percentage runs from about 10% to about 50% inside one company. Anyone quoting “our margin” for an HVAC business is quoting an average of numbers that have nothing to do with each other.
Margin and ticket move in opposite directions. Installation keeps a tenth of the money and is the most profitable line in the business, because a tenth of eleven thousand dollars beats half of seven hundred and fifty. Same trap as in roofing, visible inside a single account.
One line is red. The third one cannot be bought at market price — the market is roughly three times what that line can afford, even counting repeat work. Same company, same reputation, same crews. It fails the two-point test while its neighbours pass comfortably.
That last row is what a “no” actually looks like in practice. Not a doomed business — a healthy business with one service line it should not be advertising, and which would quietly drain the other two if it shared their budget.
What scaling actually looks like
The clearest scaling story we have is from an HVAC business — a licensed trade, capital-intensive, stable team. Which is precisely why it could be scaled at all.
Five months of scaling: 745 leads on total spend of $33,968, ending with 219 leads in July at $37 each.
Cost per lead across the same five months: $67 in March, then $49, $45, $45, and $37 in July.
Read those two lines together, because the second one is what makes the first one mean anything. The volume arrived and the price of it went down while it did — $67, $49, $45, $45, $37, never once back up. That is the opposite of what buying volume normally does to a cost per lead.
That is not what most people expect. The standard assumption is that scale makes acquisition more expensive. Over a long enough run it does, because you exhaust the cheapest available demand and start competing in more expensive auctions. But that effect is often smaller than the improvement you get from actually managing the account.
The formulation worth putting on the wall:
Anyone can buy a handful of leads cheaply at low volume. The job is keeping the math intact while spend grows five or ten times.
The part that gets left out of case studies
The first month of that scale-up went backwards.
While budget was ramping, cost per lead sat at $67 — the most expensive month of the whole run. If you were the owner watching that dashboard, you would have concluded the whole thing was going the wrong way.
Two causes, both ordinary. New campaigns were in their learning period and had no conversion history to optimize against. And the search terms hadn’t been pruned yet, so budget was flowing to queries that were never going to convert.
It took until the following month of structured weekly work to pull it down — $49 in April, then $45, $45, and $37.
Every case study presents scaling as a smooth line, and it isn’t. The dip is normal, it’s predictable, and it should be budgeted for. Which gives you two practical instructions:
Expect the spike, and decide its size in advance. Before you scale, write down how bad you’re willing to let cost per lead get, for how many weeks, before you stop. That decision is easy to make calmly in advance and nearly impossible to make sensibly in week two.
Don’t scale everything at once. New campaigns all entering learning at the same moment is what produces the worst version of the dip. Ramp one thing, let it stabilize, then ramp the next.
The channel mix at scale
Same business, once it settled:
| Campaign | Spend | Leads | Cost per lead |
|---|---|---|---|
| Residential AC repair — search | $4,748 | 122 | $39 |
| Commercial HVAC maintenance | $1,958 | 35 | $55 |
| Commercial HVAC repair | $2,624 | 37 | $71 |
| AC installation — search | $1,955 | 10.5 | $186 |
Local Services Ads sat underneath all of it, carrying about two thirds of the account’s leads.
Two things worth pulling out.
The commercial campaigns cost more than the residential ones — $55 and $71 against $39. That is an argument for separating them that has nothing to do with tidiness. On one shared budget the average sits somewhere in the middle and tells you nothing: you can’t see that you’re paying nearly twice as much for a commercial lead, so you can’t decide whether the bigger contract behind it is worth the difference. It usually is. But that should be a decision, not an accident.
The installation campaigns were the expensive mistake, and they got switched off. At $186 a lead against $39 for repair, bidding directly at people shopping for a new system meant paying five times over for the few who had already decided. The install still gets sold — on site, by a technician standing in front of a fifteen-year-old unit, to a customer who came in through a repair ad. Comparing costs across channels only works if you’re comparing the same event. Compare cost per job, or don’t compare.
Urgency is split in this trade
One more structural feature of HVAC and the other licensed trades, and it changes how the campaigns are built.
A broken air conditioner in a hot climate is an emergency. Zero patience, zero queue, and the customer is calling three companies.
A system replacement, a maintenance contract, or a new installation is a planned purchase. The customer will get several quotes over several weeks, and a backlog is fine.
Both live inside the same business, and they need separate campaigns for the same reason kitchens and decks do: different urgency, different close rate, different acceptable cost per lead, different relationship to your calendar.
The emergency campaigns should breathe with today’s availability. The replacement campaigns shouldn’t — they should run steadily and build a pipeline. Running them as one budget means the emergency work, which is time-critical and price-insensitive, competes for money with the planned work, which is neither.
The handyman end of the spectrum
At the far low-barrier end, the arithmetic changes character.
A one-person operation has a capacity limit that is brutally simple — one person’s working days — and cannot be raised by spending money. No budget produces a second pair of hands.
That has one useful consequence: for a solo operator, the capacity limit binds long before the maximum you can pay does. Which means the whole question collapses to a much simpler one: can I fill my own calendar profitably? Once it’s full, additional advertising is not an investment — it’s a subscription to disappointing people.
Where the money goes instead is at the other end: raising the value of each day. Higher-value jobs, better routing so less of the day is spent driving, and repeat customers who don’t need to be bought again.
For a business of that size, the highest-return marketing is often whatever produces work without a per-lead cost — referrals, existing customers, and search visibility. When capital is the binding constraint, the cheapest customer isn’t the one with the lowest cost per lead. It’s the one who didn’t have to be prepaid.
Which group are you in?
Four questions. They take a minute and they determine how much of the method applies to you in which order.
- Could your best employee legally do this work alone next month? If yes, you’re in the low-barrier group, whatever your trade is called.
- How many people left in the last twelve months, and how many did you hire? If those numbers are close and your headcount is flat, you’re hiring to stand still, and part of your marketing budget is really recruitment.
- If demand doubled next quarter, how long would it take to serve it? In a licensed trade the answer is a hiring timeline. In a low-barrier trade the honest answer is often “we’d lose the extra.”
- Does your work split between emergency and planned? If yes, you need at least two campaign structures, and probably two maximums.
Run your own numbers per service line through the maximums calculator — one pass per line, because as the table above shows, service and installation are not the same business.
This chapter is drawn from Win in the Spreadsheet First. The full method and the same treatment per trade are in the book.
Frequently asked questions
- Why does staff turnover belong in a marketing plan?
- Because your capacity limit is an input to every calculation about budget, and in one group of trades it's stable and in the other it isn't. In a licensed trade, full capacity means something durable: you can plan a twelve-month scale-up against it, hire ahead of it, and expect the crew you have in March to be the crew you have in October. In a low-barrier trade it's a snapshot — a third of your team may turn over inside the year, and every departure takes capacity with it, sometimes along with the customers who preferred that person. Which gives a rule: in a low-barrier trade a scale-up requires a hiring plan that starts before the campaigns do.
- Is HVAC one business or several?
- Several sharing a truck, and averaging them is expensive. Rounded figures from one company's own model: HVAC service and repair at ~$750 average job and ~50% margin supports about $4.70 a click against a market charging ~$4.50 — green, barely. HVAC installation at ~$11,000 and ~10% margin supports about $11.50 against a market at ~$7 — green, comfortably. Appliance repair as a side line at ~$300 and ~36% supports about $1.90 against a market at ~$5.30 — red. Margin runs from 10% to 50% inside one company, and margin and ticket move in opposite directions, because a tenth of eleven thousand dollars beats half of seven hundred and fifty.
- What does a real scale-up look like month by month?
- Not a smooth line. One HVAC account produced 745 leads on $33,968 over five months — finishing at 219 leads in July at $37 each. But the first month of scaling went backwards: cost per lead was $67 while new campaigns sat in their learning period with unpruned search terms, and it took another month of structured weekly work to bring back. That dip is normal and predictable, and every case study that presents scaling as a smooth line is leaving it out.
- How do I avoid the cost-per-lead spike when scaling?
- Two practical instructions. Expect the spike and decide its size in advance: before you scale, write down how bad you're willing to let cost per lead get, for how many weeks, before you stop. That decision is easy to make calmly in advance and nearly impossible to make sensibly in week two. And don't scale everything at once — new campaigns all entering learning at the same moment produces the worst version of the dip. Ramp one thing, let it stabilize, then ramp the next.
- Should commercial and residential campaigns share a budget?
- No. In the HVAC account above, commercial maintenance ran $55 per lead and commercial repair $71, against $39 for residential AC repair. On one shared budget the average sits between them and tells you nothing — you cannot see that a commercial lead costs nearly twice as much, so you cannot decide whether the bigger contract behind it is worth the difference. And watch the units when comparing channels at all: a contact is not a booked job. Compare cost per job, or don't compare.